loubnabnl/santacoder-finetuning

Issues

• Cuda OOM even with batch size 1 and lower seq length on 16GB GPU

  #16 opened a year ago by noobmldude
  3

• FIM Evaluation

  #26 opened 10 months ago by Hambaobao
  0

• How to inference the model

  #25 opened 10 months ago by athmanar
  1

• Formatting FIM data

  #22 opened a year ago by gojkoc54
  1

• Ways to reproduce this approach

  #24 opened a year ago by Yuyz0112
  2

• ValueError: Batch does not contain any data (`None`). At the end of all iterable data available before expected stop iteration.

  #23 opened a year ago by cmosguy
  2

• moving a fine tuned model to gpt_bigcode

  #20 opened a year ago by Vipitis
  2

• Can I use it to fine tune bigcode/starcoderplus?

  #21 opened a year ago by tclxmeng-jia
  1

• Question (not issue) related to dataset generation

  #14 opened a year ago by insop
  2

• Why is the inference speed using fp16 or bf16 similar to fp32?

  #19 opened 2 years ago by lionday
  1

• Has Santacoder done any pre-training in C or C++?

  #18 opened a year ago by lionday
  4

• I only generated three files after fine-tuning, is it correct? Why is there no tokenizer.json file?

  #17 opened a year ago by chen-lee-li
  1

• I used multiple GPUs for training, but there were the following errors. How can I solve them?

  #15 opened a year ago by chen-lee-li
  1

• Issue running inference on Huggingface after model upload

  #13 opened 2 years ago by umm-maybe
  2

• Linked Colab isn't working

  #12 opened 2 years ago by umm-maybe
  0

• weird behavior when setting batch_size

  #1 opened 2 years ago by DachengLi1
  2

• DeepSpeed Integration

  #5 opened 2 years ago by windspirit95
  2

• Finetuning on multiple language

  #4 opened 2 years ago by windspirit95
  2

• no_fp16 ?

  #3 opened 2 years ago by mrT23
  2